This is a second resubmission to study the response of the amygdala and orbitofrontal cortex (OFC) in patients with panic disorder (PD) to anticipatory anxiety and panic states using FOG-PET. Preclinical research has demonstrated that the amygdala is a crucial brain structure for the acquisition and expression of conditioned fear. Several different brain imaging methods have been employed to show that amygdala activation commonly occurs during the induction of fear and anxiety in humans. Some studies suggest that the threshold for amygdala activation is lower in anxiety disorder patients. Several preclinical studies have also now shown that areas within the prefrontal cortex exert control over the amygdala. Neuroimaging data also indicate this in humans and-that 1 feature of anxiety disorder patients is an alteration in activity of prefrontal cortical areas, including the anterior cingulate and the OFC. In a PET study using 15O-H20 we found that immediately prior to the administration of doxapram, which reliably produces panic attacks in PD patients but not controls, subjects who subsequently panicked after doxapram administration showed marked reduction in OFC activity. This finding indicates the need to further study the relationship between the amygdala and prefrontal cortex in PD patients. The substantial vasoconstriction caused by doxapram-induced hyperventilation made it impossible to image amygdala activation during the actual panic attack. Therefore, by adapting FDG PET to doxapram administration in a pilot study that measures metabolic rate instead of blood flow, we are able to image the amygdala and OFC during panic attacks. Our pilot data show that PD patients demonstrate increased amygdala activation more than controls during doxapram administration, but unexpectedly also show increased OFC metabolic rate. Cognitive behavioral therapy (CBT) normalized the patients' OFC response. We speculate that CBT modifies prefrontal cortical activity thereby restoring normal interaction with the amygdala. We propose to study PD patients and matched normal comparison subjects who will undergo FDG PET scans during resting, placebo, and doxapram administration. We predict that while anticipating panic, patients will show altered OFC metabolic rate compared to controls and that during doxapram administration panicking patients will show altered amygdala activation compared to non-panicking subjects. We will also look for correlations between these brain metabolic analyses and changes in autonomic, neuroendocrine, and subjective anxiety measures. Patients will then be treated with CBT and the PET scans repeated. We predict that successful treatment will lead to a normalization in OFC and doxapram-induced amygdala activity. We have added assessment of anxious anticipation and anxious arousal at the different stages of the experiment and several design changes to insure the maintenance when appropriate of an anticipatory state. The study aims to further understanding of the neuroanatomical substrate of panic attacks and to establish a possible mechanism of action for successful CBT. [unreadable] [unreadable]